Relation Between Dietary Fat Intake Type and Serum Fatty Acid Status in Children With Cystic Fibrosis

Abstract

Background and Objective

Children with cystic fibrosis (CF) and pancreatic insufficiency (PI) are at risk for fatty acid (FA) abnormalities and essential FA deficiency, with low linoleic acid (LA) and docosahexaenoic acid (DHA) concentrations and abnormal triene:tetraene (T:T) and arachidonic acid (AA):DHA ratios. The aim of the article was to determine whether type of dietary fat predicted serum LA, DHA, T:T, and AA:DHA ratios in subjects with CF and PI as compared to an unaffected comparison group.

Methods

Serum FA concentrations were assessed by capillary gas-liquid chromatography (mol%) and dietary intake by 7-day weighed food records; the 3-day coefficient of fat absorption was calculated. Total energy intake was expressed in kilocalories.

Results

A total of 65 subjects with CF and PI (8.4 ± 1.0 years, 32 girls) and 22 controls (8.5 ± 1.1 years, 13 girls) were included. Despite greater energy, saturated fat, and LA intake, the subjects with CF had lower serum LA and DHA and higher T:T and AA:DHA than those in the comparison group. Dietary total fat, monounsaturated fatty acid (MUFA), polyunsaturated fatty acid (PUFA), LA, total ω 6 polyunsaturated fatty acid (Tω6PUFA), and α-linolenic acid (ALA) intake positively predicted serum LA concentration. MUFA, total ω 3 polyunsaturated fatty acid (Tω3PUFA), and ALA intake positively predicted serum DHA concentration. Total dietary fat, MUFA, PUFA, Tω3PUFA, LA, and ALA intake negatively predicted serum T:T. ALA and Tω3PUFA intake negatively predicted serum AA:DHA.

Conclusions

Dietary fat patterns influenced serum LA, DHA, T:T, and AA:DHA in children with CF and PI. These data suggest that changes in dietary practices may result in FA profiles associated with improved clinical outcomes.

Humans lack the enzymes to desaturate fatty acids (FAs) at the 3rd and 6th carbons from the methyl end of the molecule and are dependent on dietary sources to prevent essential fatty acid deficiency (EFAD). Linoleic acid (LA; 18:2ω6) and α-linolenic acid (ALA; 18:3ω3) are the ω6 and ω3 essential fatty acids (EFAs) in humans. LA is an important membrane constituent, and has multiple cellular structural and functional roles. EFAs are precursors to the long-chain polyunsaturated FA (LCPUFA), which by themselves and by their eicosanoid products are growth factors, signal substances, inflammatory mediators, and influence gene expression (1,2). EFAD and FA abnormalities have been described in patients with cystic fibrosis (CF) for >50 years (3), and many clinical symptoms are likely influenced by these abnormalities (4–6). The most frequently described FA abnormalities are decreased LA and docosahexaenoic acid (DHA; 22:6ω3) concentrations, and abnormal triene (Mead acid; 20:3ω9 acid):tetraene (arachidonic acid [AA]; 20:4ω6) (named T:T) ratio (7–9). LA concentration was associated with growth and pulmonary status in infants and children with CF (10–13).

Increased dietary fat intake was associated with improved survival, growth, and pulmonary function (14). Although the recent CF Nutrition Consensus Reports provide guidance regarding energy intake for children with CF, specific recommendations for type of dietary fat (eg, saturated fatty acid [SFA], monounsaturated fatty acid [MUFA] or polyunsaturated fatty acid [PUFA], total ω 3 polyunsaturated fatty acid [Tω3PUFA]) to support favorable profiles were not provided (15). The aim of the present study was to determine the relationship between type of dietary fat and serum FA status in preadolescent children with CF and pancreatic insufficiency (PI). These data may provide evidence for more dietary fat recommendations for CF Care Guidelines.

METHODS

Children with CF and PI (ages 7–10 years) from 13 CF Centers participated in a study of nutritional status and progression of pulmonary disease. Cross-sectional data from the 12 months study visit are presented here. The rationale for selecting the present study time point was that dietary intake, serum FA concentrations, and anthropometric data were collected for the subjects with CF and for a contemporary comparison group of unaffected children.

The diagnoses of CF and PI were made by the home CF Center based on clinical symptoms, and duplicate quantitative pilocarpine iontophoresis sweat test, with chloride values of >60 mEq/L, and by genotype. PI was diagnosed by 3-day quantitative fecal fat collection and analysis of dietary intake, with <93% fat absorption (%CFA), and/or a stool trypsin value of <80 μg/L. Children were excluded if they had an forced expiratory volume at 1 second (FEV₁) <40% predicted, significant liver disease, insulin-dependent diabetes mellitus, Burkholderia cepacia sputum colonization, or other medical conditions or use of medications known to affect growth. Healthy children of similar age, sex, and ethnic background, who participated in a bone health study at the Children's Hospital of Philadelphia (CHOP), served as comparison subjects. These healthy children were recruited from the CHOP primary care practices and the surrounding community. Exclusion criteria were the use of any medication known to affect the growth, height or weight <3rd percentile for age and sex, percent of ideal body weight >130%, and significant developmental delay or impairments, and did not have preexisting diagnosed gastrointestinal disorders or symptoms to suggest gastrointestinal, hepatic, or pancreatic disease. Both protocols were approved by the Committee for the Protection of Human Subjects of the institutional review board at CHOP, and at the subjects’ with CF respective home institutions. Written informed consent and age-appropriate assent were obtained from the parent/legal guardian and each subject, respectively. Subjects with CF were seen at the CHOP Clinical Translational Research Center (CTRC) for overnight admissions, while in their usual state of good health. Evaluations included clinical status, anthropometry, dietary (food and supplement) intake, and phlebotomy. Subjects with CF had home fecal collections and spirometry performed at the study visit. The healthy subjects had 1 assessment performed during a 1-day study visit.

Energy Intake

Seven-day home-based weighed food records were obtained from both groups after verbal and written instructions were provided along with measuring cups, spoons, and digital food scales. Research dietitians analyzed the diet records (Nutrition Data System, Minneapolis, MN). The details of the specific brands of nutritional supplements, frequency, and dose were recorded. Total grams of fat and energy in kilocalories (kcal) from fat, carbohydrates, and protein were calculated. Percent energy expenditure requirement (%EER) was calculate for age and sex for the “active range” of physical activity as previously determined for subjects with CF (16) as well as for the healthy comparison group participants.

Intake of total fat, SFA, MUFA, and PUFA; total ω 6 polyunsaturated fatty acid (Tω6PUFA:LA + AA); Tω3PUFA (ALA + DHA + eicosapentaenoic acid [EPA; 20:5ω3]); LA; ALA; AA; and DHA were reported in both grams per day and energy intake percent (%kcal). LA and ALA intake were also expressed as a function of the adequate intake level (AI) of the dietary reference intakes (DRI) (17). The ratios of SFA:PUFA, Tω6PUFA:Tω3PUFA, and LA:ALA were all calculated (g/d).

The group with CF received detailed instructions on collection and storage for home-based 3-day stool collections. The collection was analyzed for stool total fat content by the gravimetric method (Mayo Laboratories, Rochester, MN), and the coefficient of fat absorption (%CFA) was calculated as previously described (18). This determination of %CFA was made while study subjects were on their respective habitual regimens of pancreatic enzyme replacement therapy.

Serum FA Analysis

For the subjects with CF, fasting serum phospholipid FA concentrations were determined. A nonfasting serum phospholipid FA assessment from the healthy group participants was used for comparison. In a series of studies, no significant diurnal variation was found in serum phospholipid FA status (B. Strandvik, unpublished observation). After lipid extraction according to Folch et al (19), serum phospholipids were fractionated on a single SEP-PAK aminopropyl cartridge (Waters Corp, Milford, MA), and after hydrolysis and transmethylation the FA methyl esters were separated, identified, and quantitated by capillary gas-liquid chromatography as previously described (20). All values are expressed as molar percentage (mol%) of total serum phospholipid FA.

Body Composition, Growth, and Clinical Status

Height and weight were measured using standard techniques (21) with a stadiometer accurate to 0.1 cm (Holtain, Crymych, UK), and a digital scale accurate to 0.1 kg (Scaletronix, White Plains, NY). Height was adjusted for genetic potential (22) from measured or reported biological parent heights for the subjects with CF. Z scores for height (HAZ), adjusted height (adjHAZ), weight (WAZ), and body mass index (BMI; kg/m², BMIZ) were computed (23). For the CF group, pulmonary function was evaluated by standard methods for spirometry (24,25) following inhaled albuterol and chest physiotherapy. FEV₁ percent predicted was used as the measure of pulmonary function (26,27).

Statistical Methods

Issues of sample size and power were evaluated. With a sample size of 65 subjects with CF and 22 controls, and using a 2-group t test with a 0.05 2-sided significance level, there was 80% power to detect an effect size of 0.70 between the CF and the comparison group participants. For a multiple linear regression model that already includes 4 covariates (age, sex, %CFA, %EER) with a squared multiple correlation R² of 0.25, a sample size of 65 subjects with CF had 80% power to detect, at the 0.05 level of significance, an increase in R² ≥ 0.0832 owing to the inclusion of an additional covariate (ie, a dietary intake covariate).

Histograms and 1-sample Kolmogorov-Smirnov tests were used to examine the distribution of continuous variables. Means and standard deviations (SDs) were calculated for clinical characteristics separately for subjects with CF and healthy comparison group participants, and comparisons between these 2 groups characteristics measured on a continuous scale were made using 2-group t tests (Table 1) because all of these variables had normal distributions. Comparisons between the 2 groups on categorical characteristics (ie, sex) were made using Fisher exact tests or chi-square tests. Means and SDs were tabulated by group for selected serum FA variables (Table 2), and dietary intake variables (Table 3), and comparisons between the groups were made using 2-group t tests. Log-transformed data were used for skewed variables when the transformation normalized the distribution. In cases wherein the transformation did not normalize the distribution, Mann-Whitney U tests were used.

TABLE 1.

Baseline clinical characteristics of subjects with cystic fibrosis and pancreatic insufficiency and healthy control subjects^*

	Cystic fibrosis	Controls	P
No.	65	22
Age, y	8.4 ± 1.0	8.5 ± 1.1	0.84
HAZ	–0.5 ± 1.1	–0.2 ± 0.8	0.21
adjHAZ	–0.7 ± 1.1	—	—
WAZ	–0.3 ± 1.2	–0.1 ± 0.9	0.55
BMIZ	–0.0 ± 1.1	–0.1 ± 1.0	0.76
FEV1, %	98 ± 14	—	—
CFA, %	85 ± 11	—	—
Genotype, n (%)
ΔF508/ΔF508	34 (52)	—	—
ΔF508/other	23 (36)	—	—
Other/other	8 (12)	—	—

BMIZ = body mass index z score; CFA = coefficient of fat absorption; FEV₁ = forced expiratory volume in 1 second; HAZ = height for age z score; SD = standard deviation; WAZ = weight for age z score.

^*Mean + SD; t tests used for statistical significance.

TABLE 2.

Dietary intake per day as mean (SD [standard deviation]) or median [range]^* in subjects with CF and PI and healthy control subjects

	CF, n = 65		Controls, n = 22
	Mean (median)	SD (range)	Mean (median)	SD (range)	P †
Energy intake, kcal/d	2086	417	1711	301	<0.0005
Estimated energy requirement‡, %	112	21	91	12	<0.0005
Total protein, g/d	67	16	52	11	<0.0005
Total carbohydrate, g/d	267	64	243	51	0.104
Total fat, g/d	87	21	62	12	<0.0005
%kcal from fat	37	5	32	4	<0.0005
Major fat classes
SFA, g/d	35	10	23	6	<0.0005
%kcal from SFA	15	3	12	2	<0.0005
MUFA, g/d	32	7	23	5	<0.0005
%kcal from MUFA	14	2	12	2	<0.0005
PUFA, g/d	14.2	4.4	11.4	3.7	0.007
%kcal from PUFA	6.1	1.6	6.0	1.6	0.762
Trans fats, g/d	6.2	2.2	4.8	1.3	0.007
%kcal from trans fat	2.7	0.8	2.5	0.5	0.401
Select fatty acids
Myristic acid, g/d	3.8	1.7	2.2	0.8	<0.0005
%kcal from myristic acid	1.6	0.6	1.2	0.4	<0.0005
Palmitic acid, g/d	17.1	4.7	11.3	2.6	<0.0005
%kcal from palmitic acid	7.3	1.3	6.0	1.1	<0.0005
Stearic acid, g/d	8.5	2.3	5.8	1.4	<0.0005
%kcal from stearic acid	3.6	0.6	3.1	0.6	<0.0005
OA, g/d	29.5	6.9	20.7	4.5	<0.0005
%kcal from OA	12.7	1.7	10.9	1.6	<0.0005
Palmitoleic acid, g/d	1.5	0.5	1.0	0.3	<0.0005
%kcal from palmitoleic acid	0.7	0.2	0.5	0.2	<0.0005
LA, g/d	(11.5)	(5.7–25.1)	(9.2)	(5.5–19.3)	0.005
LA, g/d	2.5	0.3	2.3	0.3	0.010
%kcal from LA	5.5	1.5	5.4	1.5	0.820
%Al§ from LA	125.3	41.3	103.1	36.0	0.027
ALA, g/d	(1.2)	(0.5–2.9)	(0.8)	(0.5–1.4)	<0.0005
%kcal from ALA	(0.5)	(0.3–1.0)	(0.4)	(0.3–0.7)	0.002
%Al§ from ALA	(122)	(59–317)	(89)	(57–139)	<0.0005
AA, mg/d	83	41	67	32	0.106
EPA, mg/d	(5)	(0–180)	(5)	(0–80)	0.72
DHA, mg/d	(20)	(0–420)	(19)	(1–100)	0.61
Tω6PUFA∥, g/d	(12.9)	(5.7–25.2)	(9.3)	(5.5–19.3)	0.005
Tω3PUFA∥, g/d	(1.2)	(0.6–3.1)	(0.9)	(0.5–1.4)	<0.0005
Dietary fat ratios, g/d
LA:ALA	10.7	2.9	12.6	3.9	0.018
AA:DHA, mg/d	4.1	2.3	4.5	3.9	0.521
Tω6PUFA:Tω3PUFA	10.3	2.8	12.0	3.2	0.024
SFA:PUFA	2.6	0.9	2.2	0.8	0.032

AA = arachidonic acid; ALA = α-linolenic acid; CF = cystic fibrosis; DHA = docosahexaenoic acid; EPA = eicosapentaenoic acid; LA = linoleic acid; MUFA = monounsaturated fatty acid; OA = oleic acid; PI = pancreatic insufficiency; PUFA = polyunsaturated fatty acid; SD = standard deviation; SFA = saturated fatty acid.

^*For skewed data, median and range values are provided. Log transformation corrected for skewness in all variables except for EPA (g/d and %kcal) in the group with CF.

^†t tests were used for normally distributed variables, whereas Mann-Whitney U tests were used for non-normally distributed variables.

^‡Estimated energy requirement, % for children in the “active” range (15).

^§Adequate intake of the dietary reference intakes, which denotes the recommended average daily intake level based on observed or experimentally determined approximations or estimates of nutrient intake for a group (or groups) of apparently healthy people that are assumed to be adequate (17).

^∥Total ω6PUFA = LA + AA.

^¶Total ω3PUFA = ALA + EPA + DHA.

TABLE 3.

Selected serum fatty acid profiles (mol%) in subjects with CF and PI and control subjects^*

	CF, n = 65		Controls, n = 22
	Mean (median)	SD (range)	Mean (median)	SD (range)	P †
SFAs
Myristic acid	0.53	0.13	0.39	0.08	<0.0005
Palmitic acid	31.24	1.78	27.87	0.87	<0.0005
Stearic acid	15.10	1.16	14.03	0.79	<0.0005
MUFAs
Palmitoleic acid	0.67	0.26	0.40	0.08	<0.0005
OA	10.80	1.34	9.39	0.74	<0.0005
PUFAs
LA	22.33	2.75	24.81	2.30	<0.0005
ALA	(0.19)	(0.12–0.41)	(0.16)	(0.10–0.39)	0.001
AA	8.02	1.68	10.90	1.27	<0.0005
EPA	0.43	0.17	0.46	0.25	0.614
DHA	1.32	0.50	2.92	1.37	<0.0005
ETA	(0.17)	(0.05–1.56)	(0.11)	(0.07–0.22)	<0.0005
Ratios
AA:DHA	6.65	1.87	4.40	1.59	<0.0005
T:T (ETA:AA)	(0.02)	(0.004–0.22)	(0.01)	(0.01–0.02)	<0.0005

AA = arachidonic acid; ALA = α-linolenic acid; CF = cystic fibrosis; DHA = docosahexaenoic acid; EPA = eicosapentaenoic acid; ETA = eicosatetraenoic acid; LA = linoleic acid; MUFA = monounsaturated fatty acid; OA = oleic acid; PI = pancreatic insufficiency; PUFA = polyunsaturated fatty acid; SD = standard deviation; SFA = saturated fatty acid; T:T = triene-to-tetraene ratio.

^*Normally distributed data presented as mean (SD) and non-normally distributed variables as median (range). Severely skewed variables were log transformed.

^†t tests were used for normally distributed variables, and Mann-Whitney U tests for non-normally distributed variables.

Differences between sexes on serum FA and dietary intake outcomes were examined using 2 group t test or Mann-Whitney U test. Pearson or Spearman correlation coefficients between types of dietary fat intake and serum FA outcomes were computed. These correlations facilitated comparison with previously published studies (28,29).

Linear regression models were used to examine the ability of various dietary intake variables to predict various serum FA outcomes, controlling for age, sex, %CFA, and kcal. A hierarchical approach was used in which age, sex, %CFA, and total energy intakes were entered at step 1, and then the particular serum FA variable or dietary intake variable of interest was entered at step 2. Change in R² (ΔR²) owing to the addition of the serum FA variables or dietary predictors to the model already containing age, sex, %CFA, and total energy intake (kcal/d) was then determined.

Linear regression models were also used to examine the ability of various dietary intake variables to predict serum FA outcomes, using data from both groups (total of subjects with CF and the comparison group participants), while controlling for age, sex, %CFA, and kcal. For the comparison group participants, %CFA was not collected, and therefore values of %CFA in the normal range (ie, between 93 and 99%) were randomly assigned for these analyses. Group membership and interactions between Group and the other predictors were examined in these models.

Despite the many multiple regression models and tests of a significant increment in the R² owing to the addition of 1 dietary predictor to a regression model already containing 4 covariates (ie, age, sex, %CFA, and energy intake), no adjustment to the level of significance was made; it was suggested that not making adjustments for multiplicity was preferable as it led to fewer errors of interpretation (30–32). Therefore, a P value of <0.05 was used to denote statistical significance, and all P values were reported to allow the reader to interpret the results (supplementary Table 1, http://links.lww.com/MPG/A120).

RESULTS

A total of 65 subjects with CF and PI (8.4 ± 1.0 years; 49% girls) and 22 healthy controls (8.5 ± 1.1 years; 59% girls) participated. There were no significant differences between the subjects with CF and the healthy controls with respect to age, sex, and growth status (Table 1). The subjects with CF had normal mean FEV₁ of 98 ± 14% of predicted, and slightly impaired fat absorption, mean %CFA of 85 ± 11.

Dietary intake data are presented in Table 2. Among the dietary intake variables, LA (g/d), ALA (mg/d), ALA (%AI and %kcal), EPA (g/d and %kcal), DHA (g/d and %kcal), Tω6PUFA (g/d), and Tω3PUFA (g/d and %kcal) had skewed distributions in the CF group. In the comparison group, dietary EPAs (g/d and %kcal) were skewed. Logarithmic transformations normalized all of these distributions except for dietary EPA (g/d and %kcal) in the CF group. Median values are presented for the skewed data.

The subjects with CF had significantly higher mean %EER, total protein, and total fat (mostly in the form of SFA) intake than the comparison group participants. Similarly, total SFA, MUFA, and PUFA intake were also significantly higher for the subjects with CF than for the comparison group participants (g/d). Median LA, ALA, Tω6PUFA, and Tω3PUFA intake (g/d) were significantly higher in the subjects with CF as compared to the comparison group participants, whereas the opposite was true for the dietary LA:ALA and Tω6PUFA:Tω3PUFA ratios, respectively. On an energy intake basis, PUFA and LA intake were not different between the 2 groups. There were no significant differences between the 2 groups for the LCPUFA intake, that is, AA, DHA, and EPA.

Major serum phospholipid FA data are presented in Table 3. There were significant differences for serum FA status between the 2 groups. Serum SFA and MUFA were higher in the subjects with CF as compared to the comparison group participants, most likely reflecting the significantly higher intake because carbohydrate intake did not differ. PUFA were lower in the subjects with CF. Serum LA, AA, and DHA concentrations were lower, whereas ALA and eicosatrienoic (mead) acid (ETA) were significantly higher in the subjects with CF. Despite greater median dietary LA intake, median serum LA concentration was lower in the subjects with CF and 40% had LA ≤21 mol%. In a report of serum PUFA status and clinical outcomes involving the subjects with CF in the present study, this LA serum concentration cutoff point was shown to be associated with improved nutritional, growth, and pulmonary outcomes (10). Only 12% of the subjects with CF had LA concentrations >26 mol% (which is the most frequently used laboratory reference cutoff point value (10)). The T:T and AA:DHA ratios were also significantly higher in the subjects with CF; depending on the T:T cutoff point used, between 17% (T:T > 0.04 (33)) and 49% (T:T > 0.02) of the subjects with CF and none of the comparison subjects had EFAD.

In subjects with CF, age was negatively correlated with serum LA concentrations (Pearson r = −0.28, P = 0.025), and positively with myristic (14:0) acid (Pearson r = 0.32, P = 0.009, and palmitic (16:0) acid (Pearson r = 0.33, P = 0.007). A marginal, nonsignificant correlation was observed between age and serum palmitoleic (16:1w9) acid (Pearson r = 0.24, P = 0.057). Similarly, the correlation between age and serum T:T ratio was inconclusive (Spearman r = 0.24 [P = 0.058] and log-transformed T:T ratio Pearson r = 0.16 [P = 0.21]). These findings may suggest that EFAD is progressive with age.

In the comparison subjects, Pearson correlation coefficients indicated that age was positively correlated with a number of dietary intake variables, including total fat (g/d) (r = 0.66, P = 0.001), SFA (g/d) (r = 0.47, P = 0.028), MUFA (g/d) (r = 0.69, P < 0.0005), and trans fats (g/d) (r = 0.54, P = 0.009). No clear correlations were observed between age and PUFA intake (g/d), LA (g/d), ALA (mg), Tω6PUFA, AA (g/d), AA (%kcal), and serum concentrations of palmitic and palmitoleic acids. There was no correlation between age and serum ALA.

In the subjects with CF, differences between the sexes were observed in the serum T:T ratio, total intakes of fat (g/d), SFA (g/d), and MUFA (g/d) in favor of the boys (P < 0.05). In the comparison group, differences between the sexes were observed in serum DHA, with girls having statistically higher concentrations (P < 0.05).

Pearson and Spearman bivariate correlation coefficients between dietary fat intake and serum PUFA concentrations in CF subjects indicated an association between the LA intake (kcal) and LA serum concentrations (r = 0.25, P = 0.048). There was also an association between DHA intake after logarithmic transformation and DHA serum concentrations using Pearson correlations, DHA_log (g/d) with DHA in serum: r = 0.29, P = 0.020 and DHA_log (kcal) with DHA in serum: r = 0.35, P = 0.004.

Total fat intake (kcal) was positively associated with the serum LA (r = 0.253, P = 0.042) and negatively with serum ALA concentrations (r = −0.266, P = 0.032). SFA intake (grams) was negatively associated with serum ALA and AA concentrations (r = −0.245, P = 0.049, and r = −0.252, P = 0.043, respectively), and positively with the log-transformed serum T:T ratio (T:T_log; r = 0.299, P = 0.015). The MUFA intake (kcal) was positively associated with serum LA and DHA concentrations (r = 0.266, P = 0.032, and r = 0.250, P = 0.045, respectively), and negatively with serum T:T_log ratio (r = −0.252, P = 0.043). The PUFA intake (kcal) was negatively associated with T:T_log ratio (r = −0.312, P = 0.011).

Regression model analyses of dietary intake predicting selected serum PUFA controlling for age, sex, kcal, and CFA% for the subjects with CF are presented in supplementary Table 1 (http://links.lww.com/MPG/A120). Serum LA concentration was positively predicted by dietary total fat (Coeff = 0.070, standard error [SE] = 0.029, R² = 0.266, ΔR² = 0.077, P = 0.017), MUFA (Coeff = 0.176, SE = 0.080, R² = 0.251, ΔR² = 0.063, P = 0.033), PUFA (Coeff = 0.232, SE = 0.091, R² = 0.271, ΔR² = 0.082, P = 0.014), LA (Coeff = 0.234, SE = 0.096, R² = 0.264, ΔR² = 0.076, P = 0.018), ALA (Coeff = 2.408, SE = 0.945, R² = 0.271 ΔR² = 0.083, P = 0.014), and Tω6PUFA (Coeff = 0.233, SE = 0.096, R² = 0.264, ΔR² = 0.076, P = 0.018) intake. Serum DHA concentrations were positively predicted by dietary MUFA (Coeff = 0.031, SE = 0.014, R² = 0.0.156, ΔR² = 0.069, P = 0.035), ALA (Coeff = 0.544, SE = 0.164, R² = 0.235, ΔR² = 0.148, P = 0.002), and Tω3PUFA_log (Coeff = 0.781, SE = 0.236, R² = 0.234, ΔR² = 0.148, P = 0.002) intake. Serum T:T status was negatively predicted by total fat (Coeff = −0.012, SE = 0.006, R² = 0.391, ΔR² = 0.046, P = 0.043), MUFA (Coeff = −0.041, SE = 0.016, R² = 0.409, ΔR² = 0.064, P = 0.016), PUFA (Coeff = −0.044, SE = 0.019, R² = 0.401, ΔR² = 0.056, P = 0.024), LA (Coeff = −0.043, SE = 0.020, R² = −0.394, ΔR² = 0.048, P = 0.037), ALA (Coeff = −0.533, SE = 0.194, R² = 0.422, ΔR² = 0.076, P = 0.008), Tω6PUFA (Coeff = −0.043, SE = 0.020, R² = 0.394 ΔR² = 0.049, P = 0.037), and Tω3PUFA_log (Coeff = −0.446, SE = 0.184, R² = 0.406, ΔR² = 0.061, P = 0.018) intake. The serum AA:DHA ratio was negatively predicted by ALA (Coeff = −2.146, SE = 0.657, R² = 0.194, ΔR² = 0.151, P = 0.002) and Tω3PUFA_log (Coeff = −2.666, SE = 0.967, R² = 0.155, ΔR² = 0.113, P = 0.008) intake.

Type of dietary fat did not predict serum ALA, and DHA intake did not predict serum DHA status (data not shown). SFA intake did not predict any of the serum FA outcomes of clinical interest. LA intake did not predict serum AA concentration (data not shown). Dietary Tω6PUFA:Tω3PUFA did not predict any of the serum outcomes of interest (data not shown).

The effects of dietary predictors in models using data from both groups were generally similar to results from models that used data from CF subjects only (data not shown), with the exception of the dietary predictor MUFA for serum outcome of DHA, which was no longer statistically significant. There was a consistent statistically significant effect of sex and a statistically significant group-by-sex interaction effect for serum DHA outcome. There were no statistically significant group-by-dietary predictor interaction effects for any of the dietary predictors at any of the serum outcomes, after controlling for confounders, indicating that the observed effects of dietary predictor were not different for the 2 groups.

DISCUSSION

Despite higher energy and LA intake in the subjects with CF, their serum LA concentration was significantly lower than in the comparison group. Dietary fat composition, when adjusting for energy intake, stool energy losses, age- and sex-predicted serum PUFA status for the subjects with CF. Type of dietary fat intake predicted serum FA outcomes in both the subjects with CF and in the comparison subjects in a similar manner. These findings have not been previously reported in preadolescent children with CF and PI.

The PUFA of clinical interest in CF are LA, DHA, T:T, and AA:DHA ratios. LA concentration was shown to be associated with growth and lung function in pediatric subjects with CF (10–13), and is considered a more relevant EFA status marker than the T:T ratio (10). ALA has 2 known biological roles: to serve as an energy substrate, and as a precursor to ω3LCPUFA-EPA, and, to a lesser extent, of DHA. ω3LCPUFA have anti-inflammatory properties and decreased the neutrophilic leukotriene (LTB4/LTB5) inflammatory ratio in subjects with CF (34). DHA constitutes the denominator in the AA:DHA ratio, which is an inflammatory index elevated in CF nasal epithelium (35). Total fat intake has been demonstrated to be associated with improved clinical outcomes (14). Given that LA and DHA are described as the most consistently decreased PUFA in CF (7,9), and that they have clinical and biochemical importance, determining dietary influences on these serum PUFA concentration are relevant to clinical care.

The relationship between different types of dietary fat intake and serum or plasma PUFA status was investigated in Italian subjects (29) with CF of similar age and sex to those in the present study and in Spanish adult subjects with CF (28). Similar to the present study, the subjects in both of these European studies had higher total fat and energy intake than control subjects. In all of these studies, serum LA and DHA status were lower in the subjects with CF as compared to the respective control groups. There were no associations between type of dietary fat intake and serum LA and DHA in either of these prior studies (28,29). Specifically, dietary LA intake has been previously reported to influence serum LA status in infants with CF (12), as well as in children of 6 to 18 years with CF (11); logistic regression and multiple regression analyses were used, respectively, in these specific studies. The relationships between other types of dietary fats and serum LA and other FA of interest, however, were not reported.

Differences between results across these studies may also reflect additional dietary differences, complexities, and multiple influences on FA metabolism in CF.

In the present study, the rationale for controlling for age and sex in the regression models relate to previous findings that serum PUFA was associated with age and sex (10). Fat and energy are related variables and provided the rationale for controlling for energy intake and losses. Fecal fat excretion has been shown to be highly correlated with fecal energy losses (36), and was therefore included in the regression analyses. By controlling for energy variables in this manner, this statistical modeling allowed us to reduce potential confounding influences to type of dietary fat.

In the present study, total fat intake was associated with serum LA and inversely with the T:T ratio by both simple correlations and in the regression models, underlying its importance in CF. LA intake predicted serum LA status and negatively predicted the T:T ratio, but did not predict serum AA (data not shown) or the AA:DHA ratio.

Dietary ALA and Tω3PUFA intake influenced serum DHA concentration, whereas DHA intake did not, probably because of the extremely low intake of DHA. The efficiency of conversion of ALA to its ω3 LCPUFA products EPA and DHA is low, approximately 5% and 0.5%, respectively (37,38). ALA is consumed in gram amounts whereas DHA and EPA are consumed in milligrams in the typical US diet (39,40). As such, dietary ALA intake may also influence serum DHA concentration in the US population. ALA and Tω3PUFA intake also negatively predicted the AA:DHA ratio (specifically, by predicting higher DHA). Both LA and ALA intake negatively predicted the serum T:T ratio, which is consistent with what is known about their metabolism (41).

ALA intake also interestingly influenced LA concentration. There are 2 likely explanations for this finding. First, although ALA is a precursor for EPA and DHA, it also serves as an energy substrate, which may spare the use of LA as an energy substrate. Second, the Δ6 desaturase (rate limiting step in the pathways from precursor PUFA to their respective LCPUFA) demonstrates substrate preferentiality for ω3 > ω6 > ω9 PUFA. This substrate preferentiality for ALA over LA may inhibit elongation of LA to its respective LCPUFA, which may also explain the positive effect of dietary ALA on serum LA concentration. Similar enzymatic substrate preferentiality and pathway interactions have been suggested by supplementation of EPA and DHA that resulted in improved LA status in subjects with CF (42). Increasing the intake of specific fats such as ALA in individuals with CF may allow for the preservation of LCPUFA by both providing a substrate for energy and influencing the preservation of LA via mechanisms described above.

Supplementation studies conducted with LA and DHA in animal models and human subjects have demonstrated that individual FA supplementation improves serum or tissue concentrations of that particular FA (42–46). These single FA supplementation studies, however, seem to demonstrate less influence overall on clinical status. Nevertheless, a recent 1-year trial of a combination of dietary fat supplements (with EPA, DHA, LA, and γ-linolenic acid) in adults with CF improved not only serum LA, DHA, and decreased AA:DHA, but also improved lean body mass, FEV₁, decreased respiratory exacerbations, days of antibiotic used, and inflammatory markers (47). Similar positive clinical results have not been obtained in the many recent ω3 LCPUFA supplementations in subjects with CF (48).

Modification of type of dietary fat intake may be a more practical and effective strategy improving serum PUFA concentrations than the use of supplements. Dietary ALA intake can be increased by consuming more flaxseed, walnut, and rapeseed or canola oils. Several brands of margarines are now frequently supplemented with flaxseed oil, ALA, and/or DHA for example, and can be encouraged as part of dietary intake. People with CF in particular may benefit from increasing intake of LA to compensate for increased turnover. LA can be derived from corn, safflower, and sunflower oils. Some food groups are rich in LA and ω3 (LC)PUFA, such as light tuna canned in oil, and may be of particular benefit to patients with CF (49).

In summary, beyond the prevention of EFA deficiency, the type of dietary fat intake influences serum FAs associated with clinically important outcomes and with inflammation. Further study is required to corroborate and expand on these findings and to determine the effect on clinical care.